Journalism without journalists: Operational and AI-generated content analysis in the first comprehensive synthetic media

Periodismo sin periodistas: Análisis operativo y de contenidos generados con IA en los primeros medios sintéticos integrales

Jornalismo sem jornalistas: Análise operacional e de conteúdos gerados por IA nos primeiros meios sintéticos integrais

César Fieiras Ceide^1*
Tania Fernández Lombao^1**
Miguel Túñez López^1***

¹ University of Santiago de Compostela (USC), Spain
^* Assistant Professor in the Department of Communication Sciences at the University of Santiago de Compostela (USC), Spain. Email: cesar.fieiras.ceide@usc.es
^** Assistant Professor in the Department of Communication Sciences at the University of Santiago de Compostela (USC), Spain. Email: tania.fernandez.lombao@usc.es
^*** Professor in the Department of Communication Sciences at the University of Santiago de Compostela (USC), Spain. Email: miguel.tunez@usc.es

Abstract

Artificial intelligence is driving a new generation of media managed primarily by algorithms, characterised by operational autonomy and minimal human oversight. This article examines the phenomenon of so-called synthetic media, analysing the operating model, impact and ethical and technological challenges of the first and only two comprehensive synthetic media: Intar Radio and NewsGPT. The research is based on in-depth interviews with the managers of both initiatives, on the qualitative analysis of their platforms, content generated and corporate communications. The results reveal an emerging ecosystem that combines technological innovation with commercial strategies aimed at specific audience niches. These media operate predominantly autonomously, using advanced AI tools for content production and distribution, although they face barriers such as the limited quality of synthetic voices, associated ethical challenges and public acceptance. The main conclusion highlights that synthetic media do not seek to replace or compete directly with traditional models, but rather to fill specific market niches, meeting specific content demands and offering complementary services that diversify the media ecosystem.

Keywords
Artificial intelligence; Synthetic media; Journalism; Automation; Innovation; Technology.

Resumen

La inteligencia artificial está impulsando una nueva generación de medios de comunicación gestionados principalmente mediante algoritmos, caracterizados por su autonomía operativa y una mínima supervisión humana. Este artículo examina el fenómeno de los llamados medios sintéticos y analiza el modelo de funcionamiento, el impacto y los desafíos éticos y tecnológicos de los dos primeros y únicos medios sintéticos integrales: Intar Radio y NewsGPT. La investigación se sustenta en entrevistas en profundidad con los directivos de ambas iniciativas, en el análisis cualitativo de sus plataformas, contenidos generados y comunicados corporativos. Los resultados evidencian un ecosistema emergente que combina innovación tecnológica con estrategias comerciales dirigidas a nichos específicos de audiencia. Estos medios operan de manera predominantemente autónoma, utilizando herramientas avanzadas de IA para la producción y distribución de contenidos, aunque enfrentan barreras como la calidad limitada de las voces sintéticas, los retos éticos asociados y la aceptación por parte del público. La conclusión principal destaca que los medios sintéticos no buscan reemplazar a los modelos tradicionales ni competir directamente con ellos, sino cubrir nichos de mercado específicos, atendiendo demandas concretas de contenido y ofreciendo servicios complementarios que diversifican el ecosistema mediático.

Palabras clave
Inteligencia artificial; Medios sintéticos; Periodismo; Automatización; Innovación; Tecnología.

Resumo

A inteligência artificial está a impulsionar uma nova geração de meios de comunicação geridos essencialmente por algoritmos, caracterizados por uma autonomia operacional e uma supervisão humana mínima. Este artigo examina o fenómeno dos chamados media sintéticos, analisando o modelo de funcionamento, o impacto e os desafios éticos e tecnológicos dos primeiros e únicos dois media sintéticos abrangentes: a Intar Radio e o NewsGPT. A investigação baseia-se em entrevistas em profundidade com os gestores de ambas as iniciativas, na análise qualitativa das suas plataformas, conteúdos gerados e comunicação corporativa. Os resultados revelam um ecossistema emergente que combina inovação tecnológica com estratégias comerciais dirigidas a nichos específicos de audiência. Estes meios de comunicação operam de forma predominantemente autónoma, utilizando ferramentas avançadas de IA para a produção e distribuição de conteúdos, embora enfrentem barreiras como a qualidade limitada das vozes sintéticas, os desafios éticos associados e a aceitação pública. A principal conclusão salienta que os meios de comunicação sintéticos não procuram substituir ou competir diretamente com os modelos tradicionais, mas sim preencher nichos de mercado específicos, satisfazendo exigências de conteúdos específicos e oferecendo serviços complementares que diversificam o ecossistema dos meios de comunicação.

Palavras-chave
Inteligência artificial; Media sintéticos; Jornalismo; Automação; Inovação; Tecnologia.

1. Introduction

Artificial intelligence has profoundly transformed the media ecosystem, introducing new dynamics in the production and distribution of news content. In this context, synthetic media are emerging, platforms managed entirely by algorithms that generate information without the intervention of journalists or technicians (Mayoral et al., 2024). Unlike other innovations, these projects do not aim to replace traditional journalism, but rather to complement it with an automated form of niche communication characterised by immediacy, scalability and personalisation.

Intar Radio and NewsGPT operate autonomously through robotisation processes (Russell and Norvig, 2010), representing an evolution in the automated generation of text, audio and video (Ufarte-Ruiz et al., 2023; Diakopoulos, 2019; Lindén, 2017). They have emerged in response to the growing demand for accessible and adaptive content, which AI facilitates by automating tasks that traditionally required human intervention.

The development of these media is the result of the convergence of emerging technologies. The constant updating of AI, Big Data, computing, robotics and blockchain tools, within the framework of Revolution 4.0 (Schwab, 2016), expands the potential of algorithmic systems in areas previously ly exclusive to human professionals. Among the most significant advances is the cloning of human voices in conversational contexts (Frąckiewicz, 2023), which transforms narrative possibilities. However, authors such as Ammerman (2020) warn of the need to legally regulate this phenomenon as part of the right to personality.

While automation is often integrated into hybrid intelligence models (Sheikh, Prins, and Schrijvers, 2023), where journalists and algorithms collaborate (Gutiérrez-Caneda, Vázquez Herrero, and López García, 2023; Mittal and Kumuraguru, 2014), synthetic media represent a different proposition: the creation of content in fully automated environments, without human editorial intervention. Far from replacing the traditional model, they offer an alternative that diversifies the media ecosystem and covers information niches through innovative solutions.

This study analyses the only two fully operational cases of synthetic media identified globally: Intar Radio and NewsGPT. Following an exploratory review of open sources and analysis of verifiable digital presence, it was found that these are the only projects managed almost exclusively by AI. Using a qualitative approach, their operating models, platforms and content are examined, assessing their impact in terms of information quality, transparency and user perception. In a context where AI is redefining news production, this study offers insights into understanding the role of synthetic media as an emerging modality that broadens the contemporary media landscape.

1.1. Origin and evolution of synthetic media

Synthetic media (Ufarte-Ruiz et al., 2023) has emerged as a result of advances in artificial intelligence (AI) and the transition from the automation of journalistic tasks to the generation of content without human intervention, marking a structural change in the global media ecosystem (Ferrucci, 2023; Gutiérrez-Caneda et al., 2023). This process began in the 2010s with automated journalism, when agencies such as the Associated Press used AI to generate financial reports, increasing the speed and scale of news production (Ufarte-Ruiz et al., 2023). At the same time, in the audiovisual field, virtual presenters such as those from Xinhua emerged, capable of simulating human language and expressions for reading the news on television (Xinhua News Agency, 2018).

The next step was the design of synthetic spaces within news agencies or television and radio schedules. Cases such as the Japanese agency JXPress or the virtual presenter of People’s Day (China) have demonstrated the viability of media managed solely by algorithms capable of collecting, writing and distributing content in real time.

In Europe, broadcasters such as Couleur 3 (Switzerland) and Expres FM (Czech Republic) have integrated synthetic voices, while in Poland, human announcers have been replaced by AI at OFF Radio Kraków, and Radio Picary has been using Basia, a virtual announcer who answers questions from the public, since 2024. The trend is spreading to Slovakia, with Europea 2, and the Czech Republic, where Hacsiko presents the morning show on Rádio Express FM. These cases fuel the debate on the impact on journalism jobs and the authenticity of radio communication (Furtáková and Janáčková, 2023).

In Spain, Hisperia, a robot on Radio 3 (RTVE), was the first national experiment aimed at Generation Z, and was a finalist in the 2023 IBC awards. It was followed by Cadena SER with the synthetic voice Victoria in Carrusel Deportivo, a precursor to the Verificaudio project (2024), aimed at detecting deepfakes. In April 2024, Radio Serranía (Cuenca) launched a programme with two synthetic announcers: VIRTUdes casAL and VIctor baRTUAL.

The challenges of synthetic media include the credibility, transparency and intellectual property of AI-generated content (Díaz-Noci, 2023). Although they allow for the optimisation of news production, they raise concerns about journalistic ethics and the relationship with the audience. The lack of editorial control can lead to errors or bias, and the absence of human empathy weakens trust in the content (Ferrucci, 2023). In response, some countries are beginning to regulate this reality: Germany approved Absolut Radio AI with human supervision and transparency requirements to ensure its legality (Fieiras-Ceide et al., 2025).

Given the emergence of these first synthetic media, and with previous studies such as that by Ufarte-Ruíz et al. (2023), or the international research by Wu et al. (2019), an in-depth analysis of the first two fully operational cases: Intar Radio and NewsGPT, becomes relevant.

1.2. Artificial Intelligence in media production and distribution

The evolution of artificial intelligence (AI) has given rise to proposals such as the use of robots that mimic the behaviour of the human brain. In the field of radio, there is already software capable of acting as artificial DJs, selecting songs according to the aesthetic profile of the station (Kuyucu, 2020).

This transfer of functions from professionals to algorithms has been defined as “automated” (Carlson, 2015; Napoli, 2014), "algorithmic" (Anderson, 2013; Diakopoulos, 2019), “computational” (Cohen et al., 2011) or “high-tech” (Salaverría, 2016). Whatever the approach, algorithms, AI and Big Data are already an active part of decision-making in media companies.

Among the tasks performed by algorithms are improving workflows (Papadimitriou, 2016), combating disinformation (Manfredi-Sánchez and Ufarte-Ruiz, 2020), content personalisation (Newman et al., 2019) and the refinement of automated news (Sandoval-Martín et al., 2019), impacting all dynamics of the journalistic environment (Fieiras-Ceide et al., 2024).

AI not only solves operational tasks (Biswal and Kulkarni, 2024), but also drives new editorial, innovative and business strategies (Canavilhas, 2022; Sánchez-García, 2023; Milosavljević and Vobič, 2019; Gutiérrez-Caneda et al., 2023; Lopezosa et al., 2023). This global transformation has favoured the emergence of synthetic media and new professional profiles that interact with algorithmic systems. Tejedor and Vila (2021) define those who master software and programming as “exo-journalists”, while other authors highlight the emerging role of AI editors and platforms (Sixto-García et al., 2021; Ufarte-Ruiz et al., 2023).

According to Sjøvaag (2024), the media will reconfigure structures and processes towards models where AI not only automates tasks but also transforms the production, distribution and interaction with content (Zamith, 2019; Sørensen, 2019; Cañedo et al., 2024).

The implementation of AI focuses particularly on information production, such as news selection (García-Orosa et al., 2023), sports, weather or finance (Canavilhas, 2023), automated tasks (Harb and Qabajeh, 2024; De Lara et al., 2022), personalisation (Risi and Pronzato, 2022) and emotional analysis of audiences (Beckett and Yaseen, 2023). According to the Reuters Institute, 67% of newsrooms already use AI to transcribe, summarise or analyse images (Newman et al., 2023).

Cases such as Radar (United Kingdom), with 30,000 automated news stories per month (Nilsson, 2019), or Associated Press, which in 2014 was already producing 1.5 billion articles without human intervention, illustrate this progress.

According to WAN-IFRA (2023), more than 50% of publishers believe their newsrooms are ready to implement AI. The main uses are text creation (54%), information search (44%), workflow improvement (43%) and text correction (43%).

In the field of audio, software allows stations to be created without human broadcasters. These are still experimental initiatives, with little journalistic reporting and minimal human voices. One of the risks is the spread of fake news through algorithms (Rodríguez-Martelo et al., 2023). In response to this, the verification of deepfakes is emerging as a key challenge for audio journalism (Calderón-González et al., 2024; Cuccovillo et al., 2023; Cano-Orón and López-Meri, 2024).

This transition towards replacing people with algorithms (Kotenidis and Veglis, 2021) generates professional uncertainty (Wölker and Powell, 2018). Although the change is currently focused on areas such as data, cybersecurity, and cloud storage, authors such as Díaz-Noci (2023) foresee the consolidation of figures such as the prompt engineer, who is responsible for guiding and optimising the results generated by AI.

1.3. Ethical challenges in AI-operated media

Hybrid intelligence raises an inevitable ethical debate (Ufarte-Ruiz et al., 2023; Pocino, 2021), centred on two main axes: the creation and detection of deepfakes (Souza and Santaella, 2021) and the inability of algorithms to apply the ethical codes of journalism (Gonçalvez et al., 2024). The increasing delegation of traditionally human tasks to automated systems (Russell and Norvig, 2010) highlights the need to establish robust oversight mechanisms.

Despite the progress of projects carried out with algorithms, authors such as Myllylahti (2025) insist that journalism always requires human intervention. This position is linked to concerns about the increase in deepfakes in the news environment (Karnouskos, 2020), one of the risks most associated with the use of AI in the creation of news stories.

New questions are also arising about authorship and rights in algorithm-generated journalistic production. Díaz-Noci (2023) cites a resolution by the US Copyright Office (February 2023), which attributed the intellectual property of the images in a comic book to the author, Kristina Kashtanova, for her selection and creative arrangement of the AI-generated elements. This case reopens the debate on whether those who formulate commands to the systems—prompt engineers—should assume authorship and responsibility for the resulting content, especially in contexts where there is no direct human intervention. Díaz-Noci (2023) suggests that exploitation rights could fall to the companies that own or contract these systems.

According to Ufarte-Ruiz et al. (2023), citing Montal and Reich (2017) and Weeks (2014), synthetic media have neglected the legal and ethical implications of using AI in journalism, from information verification to training in artificial intelligence or the detection of algorithmic biases.

Beyond authorship, additional challenges remain linked to the use of algorithms in radio: operational opacity (Mirghaderi et al., 2023) and data privacy risks (Walsh, 2023). All of this is part of the so-called “AI ethics boom” (Corrêa et al., 2023), which calls for clear regulations in the face of legal uncertainties, crystallised in principles such as the OECD AI Principles and the United Nations report Governing AI for Humanity, both from 2024.

2. Objectives

The main objective of this article is to analyse the first media outlets managed and operated predominantly by artificial intelligence, known as comprehensive synthetic media, identifying their operational characteristics, the business models that support them and their potential impact on the global media ecosystem. Through an analytical and descriptive approach, based on in-depth interviews with those responsible for both projects and the analysis of their platforms and content, the aim is to provide a comprehensive frame of reference on their relevance and challenges in the current communications environment. The specific objectives of the study are as follows:

• O1: To characterise the integrated synthetic media identified through a detailed analysis of their technologies, levels of operational autonomy, content strategies and target audiences.

• O2: To explore the business models used by these media to understand how they monetise their operations, their capacity to maintain economic sustainability, and the ethical and technological challenges associated with their functioning.

• O3: To evaluate the quality of the content generated and the operation of synthetic media, highlighting their ability to respond to the specific demands of audiences and their impact on the media ecosystem.

3. Methodology

This study adopts a qualitative and descriptive approach, with the aim of analysing in depth the operational dynamics, business models and quality of the content generated by the first synthetic media outlets managed predominantly by artificial intelligence: Intar Radio and NewsGPT. As an initial step, an exploratory review of open sources was carried out, aimed at identifying media outlets that operated entirely through artificial intelligence, without direct human intervention or with limited human intervention. This process included the systematic exploration of academic repositories and targeted searches in web browsers, in order to verify the actual and operational existence of this type of platform. The results confirmed that, although there are numerous experimental projects linked to the use of AI in newsrooms—such as automated sections, innovation labs, or isolated pilot projects within traditional media—Intar Radio and NewsGPT are the only cases detected that function as complete media outlets, with editorial structures managed almost exclusively by algorithmic systems.

The methodology for analysing both initiatives is based on two main techniques: in-depth interviews with those responsible for the projects and a detailed analysis of the platforms and content generated by both media outlets. This approach allows the phenomenon to be addressed from complementary perspectives, integrating qualitative data that offers a comprehensive view of these innovative media models.

The semi-structured interviews conducted with those responsible for these projects make it possible to study key aspects such as the technologies used, content production and distribution processes, the business models implemented, and the main challenges associated with the operation of these media outlets. This approach helps to identify shared patterns and relevant differences in the strategies adopted in each case. The intentional convenience sample is shown in Table 1 below:

The analysis of platforms and content focuses on the technological, functional and narrative dimensions of the media studied, structured in two comparative tables. The first evaluates the Intar Radio and NewsGPT websites against traditional media, considering variables such as design and usability, where Intar Radio’s intuitive navigation stands out compared to NewsGPT’s more sober interface. Accessibility is also included, assessing the availability of mobile apps and versatility of access, in which Intar Radio outperforms NewsGPT, although traditional media still offer more comprehensive solutions. In terms of content updating, the real-time automation of synthetic media is contrasted with the periodic updates of human editorial teams. In addition, interactivity—present in Intar Radio but limited in NewsGPT—and transparency are analysed, an aspect in which both declare the use of AI, unlike many traditional media, where it depends on editorial policy.

The second table compares textual content, considering factors such as informational quality, where NewsGPT excels in cross-validation, while Intar Radio prioritises entertainment and motivation. Both generate clear texts, although traditional media tend to offer more elaborate explanations. Coherence and fluency are also analysed: although NewsGPT has a solid structure, its lengthy texts can lose naturalness. In terms of accuracy, NewsGPT excels in automation, while Intar Radio, with its emotional approach, does not pursue the same rigour. Other aspects evaluated are creativity, neutrality and the use of multimodal resources, with NewsGPT standing out for its ability t te AI-generated images and videos, although traditional media remain more versatile in terms of formats.

This comparative approach allows for a rigorous characterisation of synthetic media operations and an evaluation of their impact on the media ecosystem. Triangulation applied to the platforms and content of Intar Radio and NewsGPT highlights their strengths in automation and personalisation, while revealing challenges in narrative depth, interactivity, and editorial diversity. This model contributes to a comprehensive understanding of the challenges and opportunities these media face in their consolidation within the current information landscape.

4. Results

4.1. General context and overview of comprehensive synthetic media

The emergence of synthetic media managed entirely by artificial intelligence (AI) marks a turning point in the global media ecosystem, introducing a communication model that combines operational autonomy and segmentation strategies. This article focuses on the only two cases that are fully operational worldwide: Intar Radio, a digital radio station aimed at female audiences, and NewsGPT, an automated news platform that seeks to offer objective and personalised information. Both projects aim to complement the market, without competing directly with traditional or digital media, but rather as parallel services that enrich the sector.

Intar Radio was created to fill a niche neglected by conventional radio, with motivational programming designed for women between the ages of 30 and 55. It uses AI-cloned voices and algorithms that select musical, and thematic content according to defined editorial criteria. Its “white radio” philosophy avoids controversial topics and prioritises positive messages. Although it only operates in digital format, it plans to expand to FM to consolidate its position in the market.

NewsGPT represents a commitment to global, automated news coverage, free from human bias. It is based on five advanced linguistic models and a cross-validation system, ensuring accuracy and veracity. Its unique value lies in hyper-personalisation—still in development—and the use of AI-generated text, images, and videos. Born as a technological experiment, it now has 30,000 regular readers, consolidating its initial acceptance.

Both media operate autonomously, albeit with different levels of supervision. At Intar Radio, human review ensures editorial consistency, while at NewsGPT it is limited to validating algorithmic results, relying on user participation to evaluate content. This interaction reinforces the link between technology and audience in the design of synthetic media.

The challenges reflect both the potential and limitations of AI in communication. Intar Radio needs to increase its follower base—currently 20,000 daily visits—to strengthen its sponsorship portfolio and consolidate its economic sustainability. NewsGPT, for its part, is working to consolidate its business model, supported by its solid base of 30,000 readers who value the simplicity and accuracy of its news.

Both initiatives share an optimistic view of the role of AI, highlighting its ability to fill niches and offer complementary h y services. Intar Radio and NewsGPT are not only transforming the relationship between audiences and content, but also opening up new avenues for diversifying media offerings, focusing on quality segmentation and personalisation that positions them as strategic players in an evolving ecosystem. Table 2 below summarises their main characteristics.

The following sections take an in-depth look at the operational and technological models of Intar Radio and NewsGPT, highlighting their unique content generation, personalisation and monitoring processes. The challenges identified will also be analysed, as well as the future prospects for these media, with particular attention to their complementary role in the media ecosystem and their ability to address the demands of specific audiences.

4.2. Operational models and technological differentiation

The operating models of Intar Radio and NewsGPT are based on advanced artificial intelligence (AI) technologies, which automate and personalise essential content production and distribution processes. Although both share a strong technological dependence, each adopts different approaches, tailored to their objectives and audiences.

Intar Radio automates the selection and broadcast of news, music and motivational messages. It uses algorithms that detect audience trends and preferences, and chooses music based on rhythm, tone and popularity. The content is narrated by AI-cloned voices, which, although effective, still have expressive limitations. Its white radio model avoids controversial topics and prioritises positive messages that seek an emotional connection.

NewsGPT is committed to a global model focused on factual accuracy. It automates news generation using a cross-referencing system that validates each piece of information with at least three independent sources. It also combines AI-generated text, images and videos to enrich the news experience. Although the process is automated, it includes occasional review by journalists to ensure editorial quality.

Both media outlets place personalisation at the centre of their approach, albeit with different focuses. Intar Radio tailors its programming to women between the ages of 30 and 55, curating content that reinforces emotional bonds. NewsGPT develops a hyper-personalisation system that adjusts the news to each user’s consumption profile. Table 3 below summarises the key operational processes and technologies of both projects, highlighting their similarities and differences.

Although both media rely on AI, they adopt different strategies in their relationship with audiences: Intar Radio prioritises a passive connection based on emotional content, while NewsGPT promotes active interaction through informative validation. These differences reflect how technology is reshaping not only production, but also the link with the user. Both projects face technical challenges—such as limited expressiveness in cloned voices or hyper-personalisation still in development—that require balancing automation and human supervision. Even so, they represent innovative models that redefine media consumption: Intar Radio as a new form of automated digital radio, and NewsGPT as an emerging standard in AI-generated global information.

4.3. Challenges, future projections and ethics in synthetic media

The evolution of synthetic media raises key challenges regarding the ability of artificial intelligence (AI) to operate in a diverse and ethically h te media ecosystem. Both Intar Radio and NewsGPT face challenges that condition their present and shape their future: ensuring transparency, guaranteeing the sustainability of their models, and overcoming technological limitations.

On an ethical level, total automation requires new standards. NewsGPT uses algorithms and cross-referencing to ensure accuracy, although its dependence on external sources introduces a margin of error that compromises objectivity. Intar Radio, which focuses on motivational content, reduces the risk of misinformation. Both agree on the importance of minimal but essential human supervision to maintain public trust.

From a technical standpoint, both media outlets are working to improve the user experience. Intar Radio has not yet managed to give its cloned voices natural emotions, while NewsGPT is adjusting its hyper-personalisation without weakening editorial consistency. Despite these barriers, both are moving towards models that are closer and more dynamic with their audiences.

Economically, their sustainability depends on diversifying revenue. Intar Radio focuses on sponsorships and brands related to its female niche, while NewsGPT is exploring partnerships with traditional media or subscription models, and both are considering expanding into new markets through culturally adapted offerings.

In terms of their projections, these two projects are leading trends in personalisation and innovation. Intar Radio plans to make the leap to FM radio and new formats such as interviews and motivational content. For its part, NewsGPT is committed to further integrating itself into the digital ecosystem, with immersive narratives and advanced segmentation. In addition, they can promote media literacy by facilitating critical analysis of automated content.

These media outlets show how AI can transform the relationship between audiences and content. By expanding the range of information on offer and meeting new demands, they are moving towards a more dynamic and inclusive ecosystem, where technology and creativity are redefining the future of communication.

4.4. Analysis of digital platforms and content

Analysis of the Intar Radio and NewsGPT websites reveals different strategies in terms of design, functionality and objectives. Intar Radio offers a modern aesthetic, intuitive navigation, mobile apps and interactive resources such as surveys, reinforcing its link with the audience, while NewsGPT opts for a sober design focused on clarity, with constant updates and no mobile app, standing out for its algorithmic transparency and absence of human bias.

Compared to traditional media, both prioritise automation and the use of AI, although they have limitations in direct interaction and complementary functions, where conventional models still offer more comprehensive solutions. Table 4 compares their platforms with those of traditional media, highlighting strengths, weaknesses and contributions to the media ecosystem.

Analysis of the texts from Intar Radio and NewsGPT reveals different approaches. Intar Radio offers clear and accessible content in its Magazine section, aimed at women between the ages of 30 and 55, with a good emotional connection but little depth of information, focusing on motivation and entertainment. NewsGPT, on the other hand, relies on cross-referencing to generate accurate, neutral and structured news, albeit lacking in emotional narratives or in-depth analysis.

Both projects stand out for their automation and personalisation, integrating AI-generated visual resources, but still face challenges in creativity, context and emotional connection. In contrast, traditional media, with human editorial teams, continue to surpass them in narrative richness and editorial complexity. Table 5 summarises these differences and areas for improvement.

The synthetic voice of Intar Radio’s Magazine is monotonous, with no prosodic variations or emotion, making it difficult to connect with the audience in an entertainment format that demands expressiveness. The news, aimed at women between the ages of 30 and 55, lacks thematic coherence, alternates with music without clear transitions, and rarely cites sources. They are only broadcast live, with no podcast option. In NewsGPT, although the voices are more accurate in terms of phonetics and reading, they also lack emotion and editorial hierarchy: the news items are mixed without thematic order or attribution of sources, and the AI-generated images do not always correspond to the content. The audio files are not available on demand and only the written version can be consulted on the website.

5. Conclusions

The analysis of Intar Radio and NewsGPT confirms that synthetic media represent a new information paradigm based on AI, automation and algorithmic strategies (Gutiérrez-Caneda et al., 2023). Both operate autonomously, optimising costs and time (Biswal and Kulkarni, 2024), but still show limitations in expressiveness, narrative and public acceptance, in line with studies that warn about the lack of contextualisation of automated content (Ufarte-Ruiz et al., 2023) (O1).

At the operational level, they stand out for their efficiency in selecting and personalising content (Newman et al., 2023). NewsGPT improves informational accuracy through cross-validation (Montal and Reich, 2017), while Intar Radio faces difficulties in generating immersive listening experiences due to the lack of naturalness in its voices (Kuyucu, 2020). These shortcomings reflect the need for more expressive and adaptable technologies (O1, O2).

In terms of perception and quality, the lack of emotion and uniform diction affect the connection with the audience. Intar Radio’s voice fails to create an accompanying environment (Frąckiewicz, 2023), and NewsGPT lacks editorial hierarchy, hindering the informative experience (Papadimitriou, 2016), which coincides with doubts about the credibility of algorithm-generated content (Wölker and Powell, 2018) (O2, O3).

From an ethical perspective, the absence of source attribution and human oversight raises challenges of transparency and editorial responsibility (Díaz-Noci, 2023). The proliferation of deepfakes and synthetic narratives requires clear regulatory frameworks (Calderón-González et al., 2024; Corrêa et al., 2023) (O3).

Economically, both models are in a consolidation phase. Intar Radio is exploring targeted advertising; NewsGPT is exploring subscriptions and partnerships. Their viability will depend on public acceptance and their differentiation from established media (Canavilhas, 2022; Sánchez-García, 2023; Milosavljević and Vobič, 2019) (O3).

In short, synthetic media expand the ecosystem with an automated and innovative model. However, their consolidation requires technical, ethical and narrative improvements. As Zamith (2019) and Sjøvaag (2024) warn, its success will depend on its ability to humanise interaction, raise editorial quality and ensure transparency. Its future does not lie in replacing traditional journalism, but in integrating itself as a complement within a hybrid media ecosystem.

Authors’ contributions

César Fieiras Ceide: Conceptualization; Methodology; Investigation; Formal analysis; Project management; Supervision; Visualization; Writing-original draft; Writing-review and editing.

Tania Fernández Lombao: Data curation; Investigation; Methodology; Validation; Visualization; Resources; Writing-review and editing.

Miguel Túñez López: Supervision; Methodology; Validation; Project management; Funding acquisition; Resources; Writing-review and editing. The authors have read and agree with the published version of the manuscript.

Conflicts of interest

All authors declare that they have no conflict of interest.

Sources of funding

This article is part of the activities of project PDC2023-145885-I00, funded by MCIN/AEI/ 10.13039/501100011033 and by the European Union “NextGeneration EU”/PRTR. It is also part of the activities of project PID2024-160337OB-I00 “Open public media in the age of Artificial Intelligence: public value, inclusive society and ecosystemic innovation (OpenPSM)” funded by the Ministry of Science and Innovation (Spain), AEI and Feder, EU.

References

Ammerman, Julia. (2020). El derecho a la voz propia como derecho de la personalidad [Doctoral Thesis, Universidade de Santiago de Compostela] Minerva. https://minerva.usc.es/entities/publication/c79fb93a-0ad1-4993-a360-a21a97bb8bd8

Anderson, Christopher William. (2013). Towards a sociology of computational and algorithmic journalism. New Media & Society, 15(7), 1005-1021. https://doi.org/10.1177/1461444812465137

Beckett, Charlie; & Yaseen, Mira. (2023). Generating Change. A global survey of what news organisations are doing with AI. https://shre.ink/ruQw

Biswal, Santosh Kumar; & Kulkarni, Anand J. (2024). Exploring the intersection of artificial intelligence and journalism: The emergence of a new journalistic paradigm. New York: Routledge.

Calderón-González, Daniel; Ábalos, Nieves; Bayo, Blanca; Cánovas, Pedro; Griol, David; Muñoz-Romero, Carlos; Pérez, Carmen; Vila, Pere; & Callejas, Zoraida. (2024). Deep speech synthesis and its implications for news verification: Lessons learned in the RTVE-UGR Chair. Applied Sciences, 14, 9916. https://doi.org/10.3390/app14219916

Canavilhas, Joao. (2022). Artificial intelligence and journalism: Current situation and expectations in the Portuguese sports media. Journalism and Media, 3, 510–520. https://doi.org/10.3390/journalmedia3030035

Canavilhas, Joao. (2023). Produção automática de texto jornalístico com IA: contributo para uma história. Textual & Visual Media, 17(2), 22-40. https://doi.org/10.56418/txt.17.1.2023.2

Cano-Orón, Lorena; & López-Meri, Amparo. (2024). Introducción al uso de la IA en periodismo. Guía de referencias y modos de uso. Universitat de València.

Cañedo, Azahara; Pérez-Seijo, Sara; & Rodríguez-Castro, Marta. (2024). Innovating public service media: A diagnosis of the innovation applied as an element of public value in Spain. En Daniel Barredo-Ibañez, Laura M. Castro, Araceli Espinosa, Iván Puentes-Rivera, & Paulo Carlos López-López (Eds.), Communication and Applied Technologies. Proceedings of ICOMTA 2023 (pp. 377–387). Gateway East, Singapore: Springer.

Carlson, Matt. (2015). The robotic reporter: Automated journalism and the redefinition of labor, compositional forms, and journalistic authority. Digital Journalism, 3(3), 416-431. https://doi.org/10.1080/21670811.2014.976412

Cohen, Sarah; Hamilton, James T.; & Turner, Fred. (2011). Computational journalism. Communications of the ACM, 54(10), 66-71. https://doi.org/10.1145/2001269.2001288

Corrêa, Nicholas Kluge; Galvão, Camila; Santos, James William; Del Pino, Carolina; Pinto, Edson Pontes; Barbosa, Camila, Massmann, Diogo; Mambrini, Rodrigo; Galvão, Luiza; Terem, Edmund; & de Oliveira, Nythamar. (2023). Worldwide AI ethics: A review of 200 guidelines and recommendations for AI governance. Patterns, 4(10). https://doi.org/10.1016/j.patter.2023.100857

Cuccovillo, Luca; Gerhardt, Milica; & Aichcroft, Patrick. (December 4, 2023). ‘Audio Spectrogram Transformer for Synthetic Speech Detection via Speech Formant Analysis’. IEEE International Workshop on Information Forensics and Security (WIFS), Nuremberg, Germany.

De Lara, Alicia; García-Avilés, José Alberto; & Arias-Robles, Félix. (2022). Implantación de la inteligencia artificial en los medios españoles: Análisis de las percepciones de los profesionales. Textual & Visual Media, 1(15), 1-16. https://doi.org/10.56418/txt.15.2022.001

Diakopoulos, Nicholas. (2019). Automating the news. How algorithms are rewriting the media. Cambridge, Massachusetts: Harvard University Press.

Díaz-Noci, Javier. (2023). Inteligencia artificial, noticias y medios de comunicación: Una aproximación jurídica desde la perspectiva de la propiedad intelectual al concepto y atribución de autoría. Textual & Visual Media, 17(1), 7-21. https://doi.org/10.56418/txt.17.1.2023.1

Ferrucci, Patrick. (2023). Artificial Journalism: How Algorithms are Reshaping News Production. Routledge.

Fieiras-Ceide, César; Túñez-López, José Miguel; & Fernández-Lombao, Tania. (2025). De la frecuencia al algoritmo: implementación de IA en las emisoras de radio de España [From frequency to algorithm: Implementing AI in Spanish radio stations]. Revista Latina de Comunicación Social, 83, 1-22. https://doi.org/10.4185/RLCS-2025-2457

Fieiras-Ceide, César; Vaz-Álvarez, M.; & Maroto González, Isaac. (2024). AI Implementation Strategies in the Spanish Press Media: Organizational Dynamics, Application Flows, Uses and Future Trends. Revista Trípodos, 55, pp. 1-23. https://doi.org/10.51698/tripodos.2024.55.01

Frąckiewicz, Mariusz. (2023). The role of AI in modern day radio. Último acceso: marzo 2025. Retrieved from: https://ts2.space/en/the-role-of-ai-in-modern-day-radio-broadcasting/

Furtáková, Lucia; & Janáčková, L’ubica. (2023). AI in radio: The game changer you did not bear coming. Media & Marketing Identity. Conference Proceedings from the International Scientific Conference 14th November 2023 Trnava, pp. 95-106. https://doi.org/10.34135/mmidentity-2023-09

García-Orosa, Berta; Canavilhas, Joao; & Vázquez-Herrero, Jorge (2023). Algoritmos y comunicación: Revisión sistematizada de la literatura. Comunicar, 74, 9–21. https://doi.org/10.3916/C74-2023-01

Gonçalvez, Adriana; Torre, Luísa, & Melo, Paulo Victor (2024). Inteligência artificial e algoritmos. Desafíos e oportunidades para os media. Labcom Comunicação & Artes.

Gutiérrez-Caneda, Beatriz; Vázquez-Herrero, Jorge; & López-García, Xosé. (2023). AI application in journalism: ChatGPT and the uses and risks of an emergent technology. Profesional de la Información, 32(5), 1-16. https://doi.org/10.3145/epi.2023.sep.14

Harb, Wafa’ Atieh Mohammad; & Qabajeh, Mohamed. (2024). Impact of Generative Artificial Intelligence on Journalism: Practice and Deontology. En J. Sixto-García, A. Quian, A. I. Rodríguez-Vázquez, A. Silva-Rodríguez, & X. Soengas-Pérez (Eds.), Journalism, Digital Media and the Fourth Industrial Revolution. Palgrave Macmillan. https://doi.org/10.1007/978-3-031-63153-5_18

Karnouskos, Stematis. (2020). Artificial intelligence in digital media: The era of deepfakes. IEEE Transactions on Technology and Society, 1(3), 138-147. https://doi.org/10.1109/TTS.2020.3001312

Kotenidis, Efthimis; & Veglis, Andreas. (2021). Algorithmic Journalism—Current Applications and Future Perspectives. Journalism and Media, 2(2), 244-257. https://doi.org/10.3390/journalmedia2020014

Kuyucu, Mert. (2020). Artificial intelligence in media: Radio automation systems as the first artificial intelligence application in media in the terms of “threats” and “opportunities.” En D. Akçay & E. Efe (Eds.), Dijital dönüşüm ve süreçler & digital transformation and processes (pp. 133-168). İstanbul Gelişim Üniversitesi Yayınları.

Lindén, Toomy Carl-Gustav. (2017). Algorithms for journalism: The future of news work. The Journal of Media Innovations, 4(1), 60-76. https://doi.org/10.5617/jmi.v4i1.2420

Lopezosa, Carlos; Codina, Lluís; Pont-Sorribes, Carles; & Vállez, Mari. (2023). Use of generative artificial intelligence in the training of journalists: Challenges, uses and training proposal. Profesional de la Información, 32(4), 1-12. https://doi.org/10.3145/epi.2023.jul.08

Manfredi-Sánchez, Juan Luis; & Ufarte-Ruiz, María José. (2020). Inteligencia artificial y periodismo. Revista Cidob d’Afers Internacionals, 124, 49-72. https://doi.org/10.24241/rcai.2020.124.1.49

Mayoral Sánchez, Javier; Mera Fernández, Montserrat & Morata Santos, Montserrat (2024). Capítulo 8. Integración de la inteligencia artificial en las redacciones: la experiencia de los medios de comunicación en España. Espejo De Monografías De Comunicación Social, (25), 187–209. https://doi.org/10.52495/c8.emcs.25.p108

Milosavljević, Marko; & Vobič, Igor. (2019). Human still in the loop: Editors reconsider the ideals of professional journalism through automation. Digital Journalism, 7(8), 1098-1116. https://doi.org/10.1080/21670811.2019.1601576

Mirghaderi, Leilasadat; Sziron, Monika; & Hildt, Elisabeth. (2023). Ethics and Transparency Issues in Digital Platforms: An Overview. AI, 4(4), 831-843. https://doi.org/10.3390/ai4040042

Mittal, Sudip; & Kumaraguru, Ponnurangam. (2014). Broker bots: Analyzing automated activity during high impact events on Twitter. ArXiv preprint arXiv:1406.4286. https://doi.org/10.48550/arXiv.1406.4286

Montal, Tal; & Reich, Zvi. (2017). I, robot. You, journalist. Who is the author? Authorship, bylines and full disclosure in automated journalism. Digital Journalism, 5(7), 829-849. https://doi.org/10.1080/21670811.2016.1209083

Myllylahti, Merja. (2025, 24 de febrero). As long as there is journalism, there will (and must) be a human in the loop. INMA International News Media Association. https://www.inma.org/blogs/research/post.cfm/as-long-as-there-is-journalism-there-will-and-must-be-a-human-in-the-loop?_zs=mAroR1&_zl=fVhh7

Napoli, Philip. (2014). Automated media: An institutional theory perspective on algorithmic media production and consumption. Communication Theory, 24(3), 340-360. https://doi.org/10.1111/comt.12039

Newman, Nic; Fletcher, Richard; Eddy, Kirsten; Robertson, Craig T.; & Nielsen, Rasmus Kleis. (2023). Digital News Report 2023. Retrieved from: https://reutersinstitute.politics.ox.ac.uk/digital-news-report/2023

Newman, Nic; Fletcher, Richard; Kalogeropoulos, Antonis; & Nielsen, Rasmus Kleis. (2019). Digital news report 2019. Reuters Institute for the Study of Journalism. https://reutersinstitute.politics.ox.ac.uk/sites/default/files/2019-06/DNR_2019_FINAL_0.pdf

Nilsson, Patricia. (2019, 10 de junio). El periodismo tradicional está en peligro de extinción. Milenio. https://www.milenio.com/negocios/el-periodismo-tradicional-esta-en-peligro-de-extincion

Papadimitriou, Aristea. (2016). The future of communication: Artificial intelligence and social networks. Media & Communication Studies. Mälmo University. http://www.diva-portal.org/smash/record.jsf?pid=diva2%3A1481794&dswid=5239

Pocino, Patrícia. (2021). Algorithms in the newsrooms. Challenges and recommendations for artificial intelligence with the ethical values of journalism. Consell de la Informació de Catalunya. https://fcic.periodistes.cat/wp-content/uploads/2022/03/venglishDIGITAL_ALGORITMES-A-LES-REDACCIONS_ENG-1.pdf

Risi, Elisabetta; & Pronzato, Riccardo. (2022). Algorithmic prosumers. En Emiliana Armano, Marco Briziarelli, & Elisabetta Risi (Eds.), Digital platforms and algorithmic subjectivities (pp. 149–165). University of Westminster Press. https://doi.org/10.16997/book54.l

Rodríguez-Martelo, Talía; Rúas-Araújo, José; & Maroto-González, Isaac. (2023). Innovation, digitization, and disinformation management in European regional television stations in the Circom network. Profesional de la información, 32, 1 (ene. 2023). https://doi.org/10.3145/epi.2023.ene.12

Russell, Stuart; & Norvig, Peter. (2010). Artificial Intelligence: A Modern Approach. Prentice-Hall, Upper Saddle River.

Salaverría, Ramón. (2016). Los medios de comunicación que vienen. En C. Sádaba, J. A. García-Avilés, & M. del P. Martínez-Costa (Eds.), Innovación y desarrollo de los cibermedios en España (pp. 255-263). Pamplona: Eunsa.

Sánchez-García, Pilar; Merayo-Álvarez, Noemí; Calvo-Barbero, Carla; & Diez-Gracia, Alba. (2023). Spanish technological development of artificial intelligence applied to journalism: companies and tools for documentation, production and distribution of information. Profesional de la información, 32(2), e320208. https://doi.org/10.3145/epi.2023.mar.08

Sandoval-Martín, María Teresa; La-Rosa-Barrolleta, Leonardo; Herranz-Fernández, Francisco Jaime; & Franco-Álvarez, Gullermina. (2019). “Estudio sobre la calidad de las noticias automatizadas en español”. En: XXV Congreso internacional SEP oportunidades y riesgos del periodismo hiperconectado, p. 180. ISBN: 978 84 90828724.

Schwab, Klaus. (2016). La Cuarta Revolución Industrial. Barcelona: Planeta.

Sheikh, Haroon; Prins, Corien; & Schrijvers, Erik. (2023). AI as a system technology. In Mission AI. The new system technology. Springer. https://doi.org/10.1007/978-3-031-21448-6_4

Sixto-García, José; Rodríguez-Vázquez, Ana Isabel; & López-García, Xosé. (2021). Sistemas de verificación en medios nativos digitales e implicación de la audiencia en la lucha contra la desinformación en el modelo ibérico. Revista de comunicación de la SEECI, 54, pp. 41-61. https://doi.org/10.15198/seeci.2021.54.e738

Sjøvaag, Helle. (2024). The business of news in the AI economy. AI Magazine. https://doi.org/10.1002/aaai.12172

Sørensen, Jannick Kirk. (2019). Public service media, diversity and algorithmic recommendation: Tensions between editorial principles and algorithms in European PSM Organizations. INRA. Retrieved from: https://pdfs.semanticscholar.org/0d63/5ba73cf3a0ad7a036692f9bb94242499248f.pdf

Souza, Carlos Eduardo; & Santaella, Lucía. (2021). Deepfakes na perspectiva da semiótica. TECCOGS – Revista Digital de Tecnologias Cognitivas, 23, jan./jun. p. 26-44. https://doi.org/10.23925/1984-3585.2021i23p26-44

Tejedor, Santiago; & Vila, Pere. (2021). Exo journalism: A conceptual approach to a hybrid formula between journalism and artificial intelligence. Journalism and Media, 2, 830–840. https://doi.org/10.3390/journalmedia2040048

Ufarte-Ruiz, María José; Murcia-Verdú, Francisco José; & Túñez-López, Miguel. (2023). Use of artificial intelligence in synthetic media: first newsrooms without journalists. Profesional de la información, 32(2), e320203. https://doi.org/10.3145/epi.2023.mar.03

Walsh, Toby. (2023). Will AI end privacy? How do we avoid an Orwellian future. AI & Society, 38(3), 1239-1240. https://doi.org/10.1007/s00146-022-01433-y

WAN-IFRA. World Association of Newspapers and News Publishers. (2023). World Press Trends Report 2024-2025.

Weeks, Lin. (2014). Media law and copyright implications of automated journalism. Journal of Intellectual Property and Entertainment Law, 4(1), 67-94. https://jipel.law.nyu.edu/vol-4-no-1-3-weeks

Wölker, Anja; & Powell, Tom. (2018). Algorithms in the newsroom? News readers’ perceived credibility and selection of automated journalism. Journalism, 22(1). https://doi.org/10.1177/1464884918757072

Wu, Shangyuan; Tandoc, Edson C.; & Salmon, Charles T. (2019). A field analysis of journalism in the automation age: Understanding journalistic transformations and struggles through structure and agency. Digital Journalism, 7(4), 428-446. https://doi.org/10.1080/21670811.2019.1620112

Xinhua News Agency. (2018, 8 de noviembre). World’s first AI news anchor makes “his” China debut. Xinhua. https://www.xinhuanet.com/english/2018-11/08/c_137591813.htm

Zamith, Rodrigo. (2019). Algorithms and journalism. In H. Örnebring, Y. Y. Chan, M. Carlson, S. Craft, M. Karlsson, H. Sjøvaag, & H. Wasserman (Eds.), Oxford Encyclopedia of Journalism Studies. Oxford: Oxford University Press.